Cardiovascular Surgery

Correlates of Delayed Recognition and Treatment of Acute

Type A Aortic Dissection
The International Registry of Acute Aortic Dissection (IRAD)
Kevin M. Harris, MD; Craig E. Strauss, MD, MPH; Kim A. Eagle, MD; Alan T. Hirsch, MD;
Eric M. Isselbacher, MD; Thomas T. Tsai, MD; Hadas Shiran, MD; Rossella Fattori, MD;
Arturo Evangelista, MD; Jeanna V. Cooper, MS; Daniel G. Montgomery, BS; James B. Froehlich, MD;
Christoph A. Nienaber, MD; for the International Registry of Acute Aortic Dissection
(IRAD) Investigators

Background—In acute aortic dissection, delays exist between presentation and diagnosis and, once diagnosed, definitive
treatment. This study aimed to define the variables associated with these delays.
Methods and Results—Acute aortic dissection patients enrolled in the International Registry of Acute Aortic Dissection
(IRAD) between 1996 and January 2007 were evaluated for factors contributing to delays in presentation to diagnosis
and in diagnosis to surgery. Multiple linear regression was performed to determine relative delay time ratios (DTRs) for
individual correlates. The median time from arrival at the emergency department to diagnosis was 4.3 hours (quartile
1–3, 1.5–24 hours; n⫽894 patients) and from diagnosis to surgery was 4.3 hours (quartile 1–3, 2.4 –24 hours; n⫽751).
Delays in acute aortic dissection diagnosis occurred in female patients; those with atypical symptoms that were not
abrupt or did not include chest, back, or any pain; patients with an absence of pulse deficit or hypotension; or those who
initially presented to a nontertiary care hospital (all P⬍0.05). The largest relative DTRs were for fever (DTR⫽5.11;
P⬍0.001) and transfer from nontertiary hospital (DTR⫽3.34; P⬍0.001). Delay in time from diagnosis to surgery was
associated with a history of previous cardiac surgery, presentation without abrupt or any pain, and initial presentation
to a nontertiary care hospital (all P⬍0.001). The strongest factors associated with operative delay were prolonged time
from presentation to diagnosis (DTR⫽1.35; P⬍0.001), race other than white (DTR⫽2.25; P⬍0.001), and history of
coronary artery bypass surgery (DTR⫽2.81; P⬍0.001).
Conclusions—Improved physician awareness of atypical presentations and prompt transport of acute aortic dissection
patients could reduce crucial time variables. (Circulation. 2011;124:1911-1918.)
Key Words: aorta 䡲 aortic 䡲 aneurysm 䡲 thoracic 䡲 diagnosis 䡲 imaging 䡲 surgery

A cute aortic dissection (AAD) represents a serious car-

diovascular emergency, with an associated mortality
rate of 1% to 2% per hour immediately after symptom onset
in historical untreated patients.1–3 Timely diagnosis is essen-
tial for successful management. Nevertheless, it is known that
the relative infrequency of AAD, coupled with clinical
presentations that may mimic more common problems, such
as acute coronary syndromes, can impede prompt establish-
ment of the AAD diagnosis. Significant delays may exist
between hospital arrival and definitive diagnosis and treat-
ment. The precise clinical and diagnostic factors that contrib-
ute to these delays are unknown.
Editorial see p 1902
Clinical Perspective on p 1918
The recent release of an inaugural set of guidelines for the
management of thoracic aortic disease by the American
College of Cardiology and American Heart Association Task
Force on Practice Guidelines will further increase profes-
sional awareness of ideal AAD care standards.4 Determina-

Received November 9, 2010; accepted July 29, 2011.

From the Minneapolis Heart Institute Foundation at Abbott–Northwestern Hospital, Minneapolis, MN (K.M.H., C.E.S.); Cardiovascular Division
(C.E.S., A.T.H.) and Division of Epidemiology and Community Health (A.T.H.), School of Public Health, University of Minnesota, Minneapolis;
University of Michigan, Ann Arbor (K.A.E., H.S., J.V.C., D.G.M., J.B.F.); Massachusetts General Hospital, Boston (E.M.I.); University of Colorado,
Denver (T.T.T.); University Hospital S. Orsola, Bologna, Italy (R.F.); Hospital General Universitari Vall d’Hebron, Barcelona, Spain (A.E.); and
University of Rostock, Rostock, Germany (C.A.N.).
The online-only Data Supplement is available with this article at http://circ.ahajournals.org/lookup/suppl/doi:10.1161/CIRCULATIONAHA.
110.006320/-/DC1.
Correspondence to Kevin M. Harris, MD, Minneapolis Heart Institute Foundation, 920 E 28th St, Ste 300, Minneapolis, MN 55407. E-mail
kharris@mplsheart.com
© 2011 American Heart Association, Inc.
Circulation is available at http://circ.ahajournals.org DOI: 10.1161/CIRCULATIONAHA.110.006320

1911
1912 Circulation November 1, 2011
tion of the most important factors contributing to diagnostic
and treatment delays is likely to improve the diagnostic and
decision-making process. Using the largest contemporary
AAD database, the International Registry of Acute Aortic
Dissection (IRAD),5 this report examines clinical and diag-
nostic variables in relation to time to diagnosis and time
between diagnosis and surgery in type A AAD patients.
Methods
Subjects
IRAD represents an investigational collaboration that has collected
information on unselected consecutive cases of AAD occurring at 24
aortic referral centers in 11 countries since January 1, 1996.5 Patients
with spontaneous AAD are enrolled by physicians at presentation or
identified retrospectively via queries of hospital discharge diagnosis,
surgical, and echocardiography laboratory databases. Diagnosis is
based on clinical history, imaging results, direct visualization of
surgically managed patients, and/or postmortem examination. Cases
of AAD that occur secondary to trauma are not included. Institu-
tional Review Board approval was obtained at each site, and
informed consent was provided by all participating patients. The
complete rationale, design, and methods of IRAD have been pub-
lished previously.5
Data Collection and Measures
A 290-item data collection instrument5 was used to collect compre-
hensive information on patient demographics, medical history, clin-
ical presentation, physical findings, imaging use and results, medical
and surgical management, and outcomes. Data were collected at
presentation or via physician review of records, and the coordinating
center reviewed each submission for face validity, completeness, and
clinical appropriateness. Pertinent to this study, data collection
included recordings of dates and times of symptom onset, initial
presentation (ie, hospital arrival), interhospital transfer (if applica-
ble), diagnosis, surgery, and clinical outcome.
The registry collects data on acute dissection only, ie, patients
presenting within 14 days of symptom onset. Patients with type A
AAD enrolled in IRAD from January 1, 1996, to January 29, 2007,
were included in this analysis (n⫽1204). Only surgically managed
patients with type A AAD were included in the analysis examining
factors responsible for delay in surgical repair. Patients were ex-
cluded from the analysis if presenting/diagnostic or surgical times
were not available. Thirty-four iatrogenic type A AADs were
included, but their inclusion or exclusion did not affect the overall
results. Type A dissection represents any dissection involving the
ascending aorta. Abrupt onset of pain was defined as sudden pain in
the chest, neck, or back with maximal intensity at onset. A chest
x-ray was defined as abnormal if it demonstrated a widened
mediastinum, abnormal cardiac or aortic contour, displacement or
calcification of the aortic wall, or pleural effusion. The imaging test
described is the first imaging test the patient underwent. Times
analyzed included those times from the initial emergency department
presentation to diagnosis and time from diagnosis to definitive
treatment (beginning of surgery).
Statistical Analysis
Summary statistics are presented as median time in hours with first
and third quartiles (Q1 and Q3). The Wilcoxon rank-sum test was
used for univariate comparisons. We used ␹2 analysis to compare
categorical variables when appropriate. Because time delay data had
positively skewed distributions, natural log transformations of time
delay were used to create a more normal distribution of the data. A
multiple-stage backward stepwise regression method was used as a
tool to create the final multiple-variable regression model. Candidate
variables with values of P⬍0.05 from the univariate analysis were
selected for introduction to the model. Variables were introduced to
the model in stages on the basis of general categories of variables. A
backward stepwise regression was performed at each stage, with an
adjusted P value for each variable of less than 0.10 being required to
remain eligible for reintroduction to the model along with the next
category of variables. Initial demographics and medical history
variables were first introduced, followed by variables for presenting
symptoms, followed by variables for signs of aortic dissection.
Collinearity statistics were monitored at each stage (including
variance inflation factor, condition index, and eigenvalues), and
variables suspect for collinearity were examined and removed as
deemed appropriate to resolve the issue.
The coefficients of the multiple linear regressions predicting the
natural log of delay times were used to estimate the relative delay
time ratios for those patients with the presence of an individual
variable compared with those without that variable. Those variables
with positive ␤ coefficients are associated with an increase in relative
delay time (delay time ratios ⬎1.0); those variables with negative ␤
coefficients are associated with a reduction in relative delay time
(delay time ratios ⬍1.0). On the basis of year of admission, patients
were categorized as the early IRAD group (1996 –1999; n⫽581) or
the later IRAD group (2000 –2007; n⫽623).
Results
The patient group consisted of 894 patients with time data
available for presentation to diagnosis with a median age of
62 years (Q1–Q3, 52–72 years); 294 patients (32.9%) were
female; and 653 (73.3%) were transferred from nontertiary
care facilities. Seven hundred fifty-one patients were ana-
lyzed for time from diagnosis to surgery, with a median age
of 60 years (Q1–Q3, 50 –70 years); 216 (28.8%) were female;
and 571 (76.4%) were transferred. The median time from
arrival at the emergency room to diagnosis was 4.3 hours
(Q1–Q3, 1.5–24 hours) and from diagnosis to surgery was 4.3
hours (Q1–Q3, 2.4 –24 hours).
The median time from arrival to diagnosis was 4.5 hours
(Q1–Q3, 1.6 –28.1 hours) in the early years (1996 –1999) and
4.2 hours (IQR, 1.5–24.0 hours; P⫽0.55) in the later years.
Similarly, the median time from diagnosis to surgery was 4.8
hours (Q1–Q3, 2.5–24.0 hours) in the early years and 4.0
hours (Q1–Q3, 2.3–20.0 hours; P⫽0.19) in the later years.
Patients in the more recent time interval were more likely to
be white (93% versus 88%; P⫽0.033), more likely to
undergo computed tomography (CT) imaging as the first
diagnostic test (68% versus 51%; P⬍0.001), and less likely to
undergo initial testing with transesophageal echocardiogra-
phy (28% versus 44%; P⬍0.001).
Features Associated With Diagnostic Time
Demographics/History
Table 1 outlines the demographic and historical features
associated with delay to diagnosis of aortic dissection. Fe-
male patients, those transferred from nontertiary care hospi-
tals, and patients who had undergone prior cardiac surgery
required significantly longer time intervals to establish the
diagnosis.
Signs and Symptoms
Patients with typical signs and symptoms of aortic dissection
were diagnosed sooner than those without these clinical clues
(Table 2). These symptoms included chest or back pain,
especially if described as worst ever, migrating, abrupt in
onset, or leg pain. In contrast, those with mild or no pain, or
with atypical features such as fever, required longer time
intervals for diagnostic confirmation.
 Harris et al Delays in Diagnosis and Treatment of Aortic Dissection 1913

Table 1. Median Time to Diagnosis and Surgery for Type A Acute Aortic Dissection by Demographics and Medical History
Time From Presentation to Diagnosis (n⫽894), h Time From Diagnosis to OR (n⫽751), h

Median (Q1–Q3) Median (Q1–Q3)

Characteristic Yes No P Yes No P

Demographics
Age ⱖ70 y 5.04 (1.77–28.13) 4.02 (1.50–24.00) 0.051 4.75 (2.13–22.13) 4.17 (2.39–24.00) 0.824
Female 6.40 (1.81–30.05) 3.94 (1.50–24.00) 0.001 4.64 (2.50–24.00) 4.22 (2.25–20.50) 0.105
White race 4.23 (1.52–24.00) 3.58 (1.18–27.54) 0.619 4.00 (2.33–19.25) 11.00 (4.50–47.84) ⬍0.001
Transferred from primary hospital 5.05 (2.00–28.28) 2.57 (1.00–10.78) ⬍0.001 4.83 (2.70–24.00) 3.21 (1.58–10.31) ⬍0.001
Hx of hypertension 4.25 (1.58–24.00) 4.23 (1.5–24.00) 0.534 4.50 (2.50–24.00) 4.02 (2.08–14.35) 0.154
Hx of Marfan syndrome 2.20 (1.17–12.42) 4.50 (1.51–24.00) 0.066 4.00 (2.50–29.70) 4.30 (2.35–24.00) 0.859
Prior cardiac surgery 18.25 (2.20–48.00) 4.00 (1.50–24.00) 0.001 17.80 (3.67–71.90) 4.00 (2.25–18.72) ⬍0.001
Prior aortic valve replacement 6.36 (1.63–24.00) 4.25 (1.50–24.00) 0.729 18.06 (2.88–144.00) 4.23 (2.33–24.00) 0.004
Prior mitral valve replacement 148.45 (25.75–216) 4.25 (1.50–24.00) 0.004 ... 4.33 (2.34–24.00) Not possible
Prior CABG 25.67 (4.73–80.41) 4.02 (1.50–24.00) ⬍0.001 24.00 (4.72–72.00) 4.12 (2.33–20.50) ⬍0.001
Prior catheterization/angioplasty 22.80 (3.15–49.08) 4.03 (1.51–24.00) ⬍0.001 5.82 (2.56–24.00) 4.00 (2.33–19.63) 0.099
Known aortic aneurysm 3.17 (1.39–24.25) 4.50 (1.56–24.00) 0.192 5.00 (2.09–24.00) 4.17 (2.33–22.00) 0.446
Prior aortic dissection 6.25 (1.97–24.00) 4.25 (1.50–24.00) 0.892 46.21 (3.83–143.46) 4.14 (2.33–20.10) 0.001
Hx of bicuspid aortic valve 11.00 (2.27–25.00) 4.5 (1.62–24.00) 0.376 7.12 (3.44–38.25) 4.00 (2.25–20.30) 0.025
Hx of aortic valve disease 7.50 (1.74–25.48) 4.00 (1.50–24.00) 0.136 5.23 (2.83–46.50) 4.14 (2.33–19.34) 0.023
Hx of diabetes mellitus 8.42 (2.25–51.88) 4.18 (1.50–24.00) 0.077 2.83 (1.43–4.26) 4.42 (2.42–24.00) 0.021
Hx of atherosclerosis 7.68 (2.25–47.25) 3.93 (1.50–24.00) ⬍0.001 4.96 (2.67–24.00) 4.03 (2.29–18.92) 0.044
OR indicates operating room; Q1, quartile 1; Q3, quartile 3; Hx, history of; cath, cardiac catheterization; and CABG, coronary artery bypass grafting.

Critically ill patients with hypotension, tamponade, signs
of ischemic lower extremities (as evidenced by pulse defi-
cits), coma, or altered consciousness were diagnosed sooner.
Conversely, presentation with congestive heart failure pro-
longed the diagnosis significantly.
Imaging/Testing
The initial diagnostic test and its results had incremental
impact on the time to diagnosis (Table 3). Patients with an
ECG suggestive of myocardial ischemia required more time
to establish the diagnosis of aortic dissection. Abnormalities
on the chest x-ray, such as a widened mediastinum, led to a
more rapid diagnosis, whereas normal films and pleural
effusions led to longer time intervals.
A diagnostic strategy including prompt chest CT scan as
the initial test was associated with the quickest diagnostic
times. In contrast, those whose initial test was magnetic
resonance imaging or catheter-based aortography required
significantly longer diagnostic times. If any of the tests were
interpreted as normal, times were delayed in contrast to those
with typical imaging clues, such as the presence of a double
lumen or intimal flap. The finding of aortic intramural
hematoma on any imaging test led to delays in diagnosis of
dissection.
Multiple Linear Regression of Variables Related to Delay
in Diagnosis
Transfer from an outside facility, fever, and normotension
were the greatest relative predictors of delay (Table 4). In
contrast, patients with typical symptoms (posterior pain,
abrupt pain, and worst pain ever) were diagnosed twice as
quickly as the others in this cohort.
Features Associated With Delay in Time From
Diagnosis to Surgery
Demographics
Features associated with delays to surgery include nonwhite
race and prior aortic dissection (Table 1). The delays in
nonwhites (n⫽103) were observed in the European cohort
(3.3 hours for whites versus 4.9 hours for nonwhites;
P⫽0.013), but not at North American sites (4.4 versus 5.1
hours; P⫽NS). In addition, patients transferred from nonter-
tiary care facilities and those with prior cardiac surgery had
significant delays.
Symptoms/Signs
Pain characteristics had a significant impact on the time to
surgery; those patients without typical chest pain (abrupt or
worst ever) arrived at surgery more slowly than those with these
characteristics. Critically ill patients who were hypotensive,
those in shock or tamponade, and those with ischemic limbs had
shorter times from diagnosis to surgery (Table 2).
Imaging/Testing
As shown in Table 3, patients whose initial evaluation
consisted of an aortogram or a cardiac magnetic resonance
imaging went to the operating room more slowly than those
without. Imaging results triggering more rapid arrival at
surgery include an intimal flap, a pericardial effusion, or a
periaortic hematoma.
Multiple Linear Regression of Factors Related to Time to
the Operating Room
The variables related to the greatest delay in time to surgery
include race other than white, history of coronary artery
1914 Circulation November 1, 2011

Table 2. Median Time to Diagnosis and Surgery for Type A Acute Aortic Dissection by Clinical Signs and Symptoms
Time From Presentation to Diagnosis (n⫽894), h Time From Diagnosis to OR (n⫽751), h

Median (Q1–Q3) Median (Q1–Q3)

Characteristic Yes No P Yes No P

Presenting symptoms
Abrupt onset of pain 3.50 (1.41–23.93) 24.00 (4.35–84.72) ⬍0.001 4.00 (2.25–15.40) 20.52 (4.50–96.00) ⬍0.001
Presence of any pain 4.01 (1.50–24.00) 24.00 (2.38–119.25) ⬍0.001 4.17 (2.27–19.54) 19.34 (3.45–109.44) 0.001
Chest pain 4.01 (1.50–24.00) 7.07 (1.67–45.00) 0.017 4.12 (2.25–18.00) 5.17 (2.83–27.03) 0.015
Back pain 3.03 (1.33–15.00) 6.20 (1.77–28.13) ⬍0.001 4.17 (2.34–23.88) 4.37 (2.43–20.88) 0.945
Abdominal pain 4.20 (1.64–24.00) 4.25 (1.50–24.00) 0.999 4.46 (2.35–24.00) 4.22 (2.33–21.50) 0.916
Severity of pain: mild 17.00 (1.63–57.50) 3.78 (1.50–24.00) 0.008 4.08 (2.58–40.75) 4.00 (2.17–15.90) 0.329
Severity of pain: severe or worst ever 3.78 (1.5–24) 17.0 (1.63–58) 0.008 4.00 (2.17–15.90) 4.08 (2.58–40.75) 0.329
Leg pain 2.57 (1.01–6.56) 4.61 (1.55–24.00) ⬍0.001 3.88 (2.58–9.00) 4.25 (2.25–24.00) 0.292
Syncope 4.25 (1.40–24.00) 4.11 (1.50–24.00) 0.727 3.15 (1.95–6.08) 4.59 (2.50–24.00) 0.001
Febrile 32.50 (16.76–108.00) 4.10 (1.50–24.00) 0.001 7.92 (2.09–24.94) 3.96 (2.25–19.81) 0.173
Signs
Pulse deficit 3.00 (1.23–9.75) 4.65 (1.61–27.77) ⬍0.001 3.46 (2.05–10.83) 4.65 (2.50–24.00) 0.007
Murmur of AI 4.02 (1.50–24.0) 4.50 (1.62–24.0 0.35 3.87 (2.5–16.50) 4.83 (2.13–24.0) 0.236
Presenting with hypotension 2.50 (1.27–7.01) 4.58 (1.60–24.25) ⬍0.001 3.50 (2.17–7.00) 4.42 (2.33–24.00) 0.066
Presenting with shock 4.41 (1.56–24) 2.67 (1.33–20.25) 0.155 2.02 (2.92–5.42) 4.42 (2.43–24.00) 0.006
Presenting with cardiac tamponade 1.71 (0.69–5.86) 4.35 (1.58–24.00) 0.002 2.54 (1.31–6.71) 4.26 (2.42–21.69) 0.004
Ischemic lower extremity 3.00 (1.04–7.81) 4.57 (1.62–24.25) 0.002 3.46 (2.25–6.18) 4.58 (2.50–24.00) 0.038
Coma/altered consciousness 2.82 (1.26–22.63) 4.75 (1.62–24.25) 0.007 3.00 (2.17–5.13) 4.67 (2.50–24.00) 0.003
CHF as sign or symptom 23.28 (1.56–65.10) 4.07 (1.50–24.00) 0.005 8.12 (2.95–24.00) 4.25 (2.34–21.43) 0.062
Admission SBP ⱖ105 mm Hg 5.05 (1.83–26.17) 2.50 (1.17–8.94) ⬍0.001 4.75 (2.50–24.00) 3.33 (2.07–6.92) ⬍0.001
OR indicates operating room; Q1, quartile 1; Q3, quartile 3; AI, aortic insufficiency; CHF, congestive heart failure; and SBP, systolic blood pressure.

bypass surgery, and the time from presentation to diagnosis
(Table 5). Patients with shock, cardiac tamponade, and
diabetes mellitus went to surgery far more rapidly than others
in the cohort.
Discussion
Significant delays exist between emergency department ar-
rival and establishment of a definitive diagnosis of AAD.
Variables associated with delayed diagnosis include female
sex, the absence of typical historical features (abrupt-onset,
chest, or back pain), or high-risk examination findings,
including hypotension or pulse deficits. Patients initially
presenting to nontertiary hospitals experienced delays in both
recognition and surgery. Furthermore, delays from diagnosis
to surgery occur more commonly in nonwhite patients, those
with a history of prior cardiac surgery, and those in whom the
initial presentation is not complicated by typical high-risk
features such as shock or tamponade.
Delays in Recognition
Prior analysis of the IRAD registry suggested that certain
variables are associated with delays between symptom onset
and diagnosis, including lack of pain, female sex, and
abdominal pain.6 – 8 To the best of our knowledge, the reasons
for delay in the time to diagnosis of aortic dissection have
previously been evaluated in only 1 small study (pleural
effusion, troponin positivity, acute coronary syndrome like
ECG and dyspnea),9 and the factors associated with delay in
operative repair have not previously been evaluated. Prior
data from tertiary referral centers suggested that in patients
with proven aortic dissections, a diagnosis other than dissec-
tion was first considered in more than one third of the
cases,10,11 and dissection was not discovered until postmor-
tem examination in nearly one third of cases.10 Additionally,
in up to two thirds of patients, ⬎1 test is necessary to
establish the diagnosis.12 As outlined in the American College
of Cardiology/American Heart Association thoracic aortic
disease guidelines, high-risk historical features or conditions
(connective tissue disorders, Marfan syndrome, Loeys-Dietz
syndrome, etc) or physical examination findings suggestive
of aortic dissection should be sought in all patients presenting
with symptoms that could represent AAD.4 In addition to
standard testing such as ECG and chest radiography, defini-
tive diagnostic imaging (CT, transesophageal echocardiogra-
phy, or magnetic resonance imaging) should be pursued in
patients at high risk for the disease based on the screening
process.4 Recognition of features associated with both typical
and atypical presentations of dissection is instructional for
improving rapid detection.
Typical Clinical Features Associated
With Dissection
Classic features suggestive of acute dissection include the
abrupt onset of severe chest pain, often of a tearing or ripping
 Harris et al Delays in Diagnosis and Treatment of Aortic Dissection 1915

Table 3. Median Time to Diagnosis and Surgery for Type A Acute Aortic Dissection by Diagnostic Testing
Time From Presentation to Diagnosis (n⫽894), h Time From Diagnosis to OR (n⫽751), h

Median (Q1–Q3) Median (Q1–Q3)

Characteristic Yes No P Yes No P

ECG
ECG normal 4.38 (1.33–28.33) 4.50 (1.67–24.0) 0.44 4.3333 (2.00–18.72) 4.1167 (2.50–24.00) 0.750
ECG with infarct, new Q waves, or 23.80 (2.45–46.25) 4.08 (1.51–24.00) 0.02 3.21 (1.75–19.69) 4.25 (2.44–20.50) 0.245
ST-segment elevation
ECG nonspecific ST-T changes 5.00 (2.00–25.57) 3.97 (1.40–24.00) 0.005 4.83 (2.50–24.00) 4.03 (2.33–18.68) 0.233
Preoperative myocardial ischemia 7.42 (1.67–28.1) 4.01 (1.55–24.00) 0.96 3.03 (1.75–5.50) 4.58 (2.50–24.00) 0.004
CXR findings
Normal 7.42 (2.25–40.79) 4.08 (1.50–24.00) 0.009 4.82 (2.03–22.69) 4.54 (2.58–24.00) 0.738
Widening of mediastinum 3.50 (1.59–24.00) 7.95 (2.00–48.00) ⬍0.001 4.25 (2.58–18.96) 5.07 (2.50–24.00) 0.203
Abnormal CXR without pain 3.00 (1.46–15.00) 4.50 (1.50–24.00) 0.02 3.42 (2.29–6.75) 4.75 (2.27–24.00) 0.035
Pleural effusion 7.05 (1.87–70.5) 4.23 (1.51–24.00) 0.01 5.17 (2.17–24.00) 4.50 (2.50–20.51) 0.213
First imaging test
CT 3.93 (1.50–24.00) 5.00 (1.61–30.50) 0.005 4.75 (2.58–21.74) 3.75 (2.00–22.75) 0.03
TEE/TTE 4.62 (1.50–29.00) 4.00 (1.50–24.00) 0.14 3.33 (2.00–18.17) 4.78 (2.67–24.00) 0.002
MRI 96.00 (25.97–156.00) 4.07 (1.50–24.00) 0.012 516.19 (72.10–996.00) 4.17 (2.33–20.50) 0.002
Aortogram 16.50 (3.11–28.30) 4.00 (1.50–24.00) 0.014 7.23 (3.19–21.40) 4.17 (2.33–22.06) 0.33
Imaging result
Normal image 8.92 (3.89–74.41) 4.10 (1.50–24.00) 0.042 4.94 (1.68–19.13) 4.33 (2.37–24.00) 0.627
Any test showed double lumen 3.50 (1.50–24.00) 6.32 (2.06–28.01) ⬍0.001 4.06 (2.38–19.00) 4.88 (2.08–24.00) 0.302
Any test showed partial 8.17 (1.77–45.38) 3.50 (1.33–24.00) 0.004 6.97 (3.27–63.95) 4.00 (2.25–13.28) ⬍0.001
false lumen thrombosis
Any test showed pericardial 4.16 (1.50–24.00) 4.05 (1.50–24.00) 0.842 3.57 (2.00–15.00) 5.00 (2.66–24.00) 0.001
effusion
Any test showed intramural 6.5 (1.9–48.0) 4.0 (1.5–24.0) 0.007 5.4 (2.1–24.0) 4.2 (2.4–20.0) 0.051
hematoma
OR indicates operating room; Q1, quartile 1; Q3, quartile 3; CXR, chest x-ray; CT, computed tomography; TTE, transthoracic echocardiography; TEE, transesophageal
echocardiography; and MRI, magnetic resonance imaging.

character, with radiation to the back and physical examination
findings such as pulse deficits and a murmur suggestive of
aortic insufficiency.13 However, in reality, although chest
pain of severe nature and its abrupt onset are present in more
than three quarters of patients with type A AAD, the
remaining 25% do not have typical features.5,13 Our data
provide supportive evidence that emergency department cli-
nicians are more likely to establish the diagnosis when typical
clinical characteristics of the pain (abrupt, posterior, and
worst ever) are present. In patients with life-threatening limb
ischemia or shock, the diagnosis of aortic dissection is
achieved more quickly because multiple diagnostic tests are
obtained concurrently. Our finding that hypotension is asso-
ciated with a more rapid diagnosis is consistent with an
earlier study.9 Once diagnosed, patients with typical symp-
toms, pulse deficits, cardiac tamponade, or shock undergo

Table 4. Time From Presentation to Diagnosis (Multiple Table 5. Time From Diagnosis to Surgery (Multiple
Linear Regression) Linear Regression)
Delay 95% CI for Delay 95% CI for
Time Delay Time Time Delay Time
B P Ratio Ratio B P Ratio Ratio
Transferred from primary hospital 1.206 ⬍0.001 3.34 2.38 to 4.69 Log normal of hours from 0.342 ⬍0.001 1.35 0.235 to 0.371
Female sex 0.550 0.001 1.73 1.27 to 2.36 presentation to diagnosis
Chest pain, posterior ⫺0.500 0.001 0.61 0.45 to 0.81 White race vs other 0.150 ⬍0.001 2.25 0.402 to 1.223
Worst pain ever ⫺0.632 0.001 0.53 0.36 to 0.78 History of diabetes mellitus ⫺0.113 0.003 0.42 ⫺1.459 to ⫺0.293
Febrile 1.631 ⬍0.001 5.11 2.07 to 12.62 History of CABG 0.139 ⬍0.001 2.81 0.469 to 1.600
Abrupt onset of pain ⫺0.848 0.002 0.43 0.25 to 0.73 Presenting shock ⫺0.083 0.031 0.63 ⫺0.887 to ⫺0.042
Admission SBP ⱖ105 mm Hg 0.895 ⬍0.001 2.45 1.80 to 3.33 Presenting tamponade ⫺0.090 0.020 0.48 ⫺1.348 to ⫺0.117
CI indicates confidence interval; SBP, systolic blood pressure. CI indicates confidence interval; CABG, coronary artery bypass grafting.
1916 Circulation November 1, 2011
surgery faster than their counterparts. The wide variation in
presenting symptoms between individuals contributes to phy-
sicians’ difficulty in establishing this diagnosis. Interestingly,
an aortic aneurysm, Marfan syndrome, or the presence of an
aortic insufficiency murmur, each of which may offer clues to
the diagnosis of dissection, was associated with earlier
diagnosis but failed to reach statistical significance.
Atypical Presentations and Contributors to
Delayed Recognition
Our data highlight several groups of patients with a delayed
diagnosis who deserve emphasis. First, as a group, women are
diagnosed more slowly. The recognition of delayed diagnosis
of dissection in women has previously been reported and was
attributed to atypical presenting symptoms.7 Women are at
particularly high risk for mortality from dissection; thus,
earlier recognition is paramount. Second, patients with heart
failure, perhaps related to concomitant aortic insufficiency,
can lead clinicians down other diagnostic and treatment
algorithms, thereby delaying the true diagnosis. Third, a fever
at presentation, not a common symptom of dissection, may
lead to alternative diagnostic strategies.5,13
Other groups experience delay in recognition and treat-
ment. Patients without pain of any type are typically not
diagnosed or treated as quickly. In fact, prior analysis
suggests that patients with painless dissection have a height-
ened risk.6 Patients with prior cardiac surgery have delayed
diagnosis and treatment.14 In fact, although 1 in 6 patients
with aortic dissection has had prior cardiac surgery, it may
not be widely recognized that this constitutes a risk factor for
dissection.14 Patients with prior surgery may have abnormal-
ities of their aortic wall from aortic cross-clamping, aortic
valve implantation, or insertion of proximal bypass grafts that
predispose them to recurrent dissection. In addition, if they
had bicuspid aortic valve disease, they may have an inherited
tendency to proximal aortic aneurysm and/or dissection.
These patients may present with atypical symptoms. Patients
initially presenting to nontertiary hospitals experience delays
in both recognition and treatment. Once recognized, interho-
spital transfer adds time and requires coordination between
hospitals.15 Nonwhite patients experience delays in surgical
treatment for reasons that are uncertain. They may be related
to access to care, but this merits further investigation. The
delays in the nonwhites were seen primarily in Europe and
may be related to language barriers, ie, not speaking the
native language of the European countries.
Diagnostic Testing
The incidence of AAD is far lower than that of acute coronary
syndromes.5,16,17 Furthermore, the ECG, which is routinely
performed when patients present with chest pain, suggests an
acute coronary syndrome in more than one fourth of patients
with acute aortic syndromes.18 As corroborated by this
analysis, an abnormal ECG leads to delays in diagnosis.9 In
AAD, the abnormal ECG may be the result of hypotension,
compromise of coronary ostia, or preexisting coronary dis-
ease.19 Although preliminary evidence suggests a possible
diagnostic role for plasma biomarkers, aside from the
D-dimer assay, these tests are not yet available in emergency
departments.20
Chest x-ray is standard in patients presenting with chest
pain. Our findings suggest, however, that clinicians may be
falsely reassured by a normal chest x-ray, and the diagnosis of
dissection is delayed in this group. Although typical radio-
graphic findings of type A AAD include widening of the
mediastinum or aortic knob, in fact, ⬎20% of patients with
confirmed AAD lack abnormalities of the mediastinum or
aortic contour.5,13 The relatively nonspecific finding of a
pleural effusion was also seen in those with longer diagnostic
times.9
The development of an evidence-based strategy to exclude
aortic dissection in patients presenting with chest pain would
be useful because ⬎4.6 million Americans annually present
to emergency departments with chest pain. A strategy of a
low threshold for CT scanning would need to be balanced by
the potential harm resulting from long-term effects of radia-
tion and contrast-induced nephrotoxicity. Rapid diagnosis of
dissection is most likely when CT or echocardiography is part
of the diagnostic testing. In contrast, when magnetic reso-
nance imaging or an aortogram was performed, the diagnosis
was delayed, often significantly. In our series, the use of
magnetic resonance images, which are not widely available
especially on an emergent basis, likely represents cases in
which an alternative diagnosis was initially contemplated.3
Patients with aortograms as part of the testing regimen may
represent patients who are being evaluated for an acute
coronary syndrome that may be complicating a dissection.
Patients with normal initial studies are less likely to have a
quick diagnosis of dissection, and physicians should realize
that tomographic imaging is excellent but not 100% sensitive
or specific. If there is a high index of suspicion, a second test
may be needed in individual cases.12,21 Intramural hematoma,
an important type of aortic dissection, made up 8.2% (n⫽72)
of this data set. These patients may not always be recognized
as having a variant of aortic dissection requiring similar
treatment and thus not surprisingly were associated with
delays in both recognition and treatment.22,23
Delays to the Operating Room
As outlined above, our study shows that some patient groups,
such as patients who transfer from other hospitals, those
without pain, nonwhites, and those with prior cardiac surgery,
are prone to delays in treatment. In addition, those with
delays in presentation to diagnosis also had delays in the time
to surgery. Those centers with delays in diagnosis and surgery
likely represent nontertiary hospitals that have overall less
familiarity with dissection. Patients with prior surgery are
often managed medically, perhaps because of a perceived
excessive risk of surgery.14 However, why other features are
associated with longer delays is not well understood, but may
be a reflection of the surgeon’s individual preferences.24 That
an earlier diagnosis of AAD and initiation of definitive
treatment are associated with improved survival has never
been definitively proven, although a systematic rapid opera-
tive approach is associated with excellent outcomes.25 It has
been postulated that the delay in recognition may partially
explain the increased mortality in certain subgroups of
 Harris et al Delays in Diagnosis and Treatment of Aortic Dissection
patients with dissection, including women and those without
pain.6,7 Although it is logical to presume that, on average,
more prolonged aortic dissection would be associated with
major end-organ compromise (eg, coronary, cerebrovascular,
spinal, mesenteric, renal, or lower-extremity ischemia; wors-
ened valvular insufficiency; and heart failure), this assump-
tion is based on observational data only.
For a condition with a mortality rate of 1%/h, a median
time to diagnosis of 4.3 hours plus an additional 4.3 hours
from diagnosis to surgery indicates that there is a significant
opportunity for systematic improvement. Little is known
from this data set about the type or the experience of the
physician first encountering the patient. The physician’s
practice experience, particularly related to aortic emergen-
cies, may be particularly relevant. Educational efforts to
improve awareness and ultimately to improve aortic dissec-
tion recognition and treatment may be relevant, particularly in
nontertiary hospitals with low exposure to aortic emergen-
cies.15 These data show a trend toward improvement with
shorter times to diagnosis and surgery in the latter times
analyzed, which may suggest increased awareness of the
condition at IRAD centers. Although this database was not
designed to measure physician awareness of dissection, these
data show that practice patterns have changed at IRAD
centers over time (ie, more frequent transfer and initial testing
with CT), and it is conceivable that an emphasis on dissection
diagnosis and treatment has beneficially affected care. A
systematic approach to the diagnosis and management of
AAD will likely reduce avoidable delays.15,25 This should
include creation of regional networks in which defined
protocols allow the most expedient diagnosis and transfer of
patients with AAD to Aortic Centers of Excellence for
definitive treatment.
Clinical Implications
These findings suggest that time to definitive diagnosis and
treatment could be improved by physician awareness of
atypical presentations of aortic dissection, a low threshold for
tomographic imaging, and prompt transfer to tertiary care.
Acknowledgments
A complete list of the IRAD investigators appears in the online-only
Data Supplement. Dr Harris had full access to all study data and
takes responsibility for integrity of data and accuracy of analysis.
Sources of Funding
IRAD is supported by grants from the University of Michigan Health
System, Varbedian Fund for Aortic Research, Hewlett Foundation,
Mardigian Foundation, and GORE, Inc.
Disclosures
Dr Eagle has received grant/research support from Bristol Myers
Squibb, Blue Cross/Blue Shield of Michigan, GORE, Hewlett
Foundation, Mardigian Foundation, Pfizer, SanofiAventis, and
Varbedian Fund and consultant fees from the National Institutes of
Health National Heart, Lung, and Blood Institute, Pfizer, Sanofi-
Aventis, and Robert Wood Johnson Foundation. Dr Froehlich has
consulted with Pfizer and Sanofi-aventis; served on the speakers ’
bureau for Pfizer, Sanofi-aventis, and Merck; and has contracted
research with Blue Cross/Blue Shield of Michigan, Mardigian
Foundation, and Fibromuscular Disease Society of America. Dr
Nienaber has received grant/research support from Varbedian Fund,
1917
Hewlett Foundation, Mardigian Foundation, and GORE, Inc. The
sponsors had no role in the design and conduct of the study; the
collection, management, analysis, or interpretation of the data; or
the preparation, review, or approval of the manuscript.
References
1. Hirst AE, Johns VJ, Krime SW Jr. Dissecting aneurysm of the aorta: a
review of 505 cases. Medicine. 1958;37:217–279.
2. Meszaros I, Morocz J, Szlavi J, Schmidt J, Tornoci L, Nagy L, Szep L.
Epidemiology and clinicopathology of aortic dissection. Chest. 2000;117:
1271–1278.
3. Tsai TT, Nienaber CA, Eagle KA. Acute aortic syndromes. Circulation.
2005;112:3802–3813.
4. Hiratzka LF, Bakris GL, Beckman JA, Bersin RM, Carr VF, Casey DE Jr,
Eagle KA, Hermann LK, Isselbacher EM, Kazerooni EA, Kouchoukos
NT, Lytle BW, Milewicz DM, Reich DL, Sen S, Shinn JA, Svensson LG,
Williams DM. 2010 ACCF/AHA/AATS/ACR/ASA/SCA/SCAI/SIR/
STS/SVM guidelines for the diagnosis and management of patients with
thoracic aortic disease: a report of the American College of Cardiology
Foundation/American Heart Association Task Force on Practice
Guidelines, American Association for Thoracic Surgery, American
College of Radiology, American Stroke Association, Society of Car-
diovascular Anesthesiologists, Society for Cardiovascular
Angiography and Interventions, Society of Interventional Radiology,
Society of Thoracic Surgeons, and Society for Vascular Medicine.
Circulation. 2010;121:e266 – e369.
5. Hagan PG, Nienaber CA, Isselbacher EM, Bruckman D, Karavite DJ,
Russman PL, Evangelista A, Fattori R, Suzuki T, Oh JK, Moore AG,
Malouf JF, Pape LA, Gaca C, Sechtem U, Lenferink S, Deutsch HJ,
Diedrichs H, Marcos y Robles J, Llovet A, Gilon D, Das SK, Armstrong
WF, Deeb GM, Eagle KA. The International Registry of Acute Aortic
Dissection (IRAD): new insights into an old disease. JAMA. 2000;283:
897–903.
6. Park SW, Hutchison S, Mehta RH, Isselbacher EM, Cooper JV, Fang J,
Evangelista A, Llovet A, Nienaber CA, Suzuki T, Pape LA, Eagle KA,
Oh JK. Association of painless acute aortic dissection with increased
mortality. Mayo Clin Proc. 2004;79:1252–1257.
7. Nienaber CA, Fattori R, Mehta RH, Richartz BM, Evangelista A, Petzsch
M, Cooper JV, Januzzi JL, Ince H, Sechtem U, Bossone E, Fang J, Smith
DE, Isselbacher EM, Pape LA, Eagle KA; International Registry of Acute
Aortic Dissection. Gender-related differences in acute aortic dissection.
Circulation. 2004;109:3014 –3021.
8. Upchurch GR Jr, Nienaber C, Fattori R, Evangelista A, Oh J, Cooper JV,
Isselbacher E, Suzuki T, Eagle KA; IRAD Investigators. Acute aortic
dissection presenting with primarily abdominal pain: a rare manifestation
of a deadly disease. Ann Vasc Surg. 2005;19:367–373.
9. Rapezzi C, Longhi S, Graziosi M, Biagini E, Terzi F, Cooke RM, Quarta
C, Sangiorgi D, Ciliberti P, Di Pasquale G, Branzi A. Risk factors for
diagnostic delay in acute aortic dissection. Am J Cardiol. 2008;102:
1399 –1406.
10. Spittell PC, Spittell JA Jr, Joyce JW, Tajik AJ, Edwards WD, Schaff HV,
Stanson AW. Clinical features and differential diagnosis of aortic dis-
section: experience with 236 cases (1980 through 1990). Mayo Clin Proc.
1993;68:642– 651.
11. Hansen MS, Nogareda GJ, Hutchison SJ. Frequency of and inappropriate
treatment of misdiagnosis of acute aortic dissection. Am J Cardiol. 2007;
99:852– 856.
12. Moore AG, Eagle KA, Bruckman D, Moon BS, Malouf JF, Fattori R,
Evangelista A, Isselbacher EA, Suzuki T, Nienaber CA, Gilon D, Oh JK.
Choice of computed tomography, transesophageal echocardiography,
magnetic resonance imaging and aortography in acute aortic dissection:
International Registry of Acute Aortic Dissection. Am J Cardiol. 2002;
89:1235–1238.
13. Klompas M. Does this patient have an acute thoracic aortic dissection?
JAMA. 2002;287:2262–2272.
14. Collins JS, Evangelista A, Nienaber CA, Bossone E, Fang J, Cooper JV,
Smith DE, O’Gara PT, Myrmel T, Gilon D, Isselbacher EM, Penn M,
Pape LA, Eagle KA, Mehta RH; International Registry of Acute Aortic
Dissection (IRAD). Differences in clinical presentation, management, and
outcomes of acute type A aortic dissection in patients with and without
previous cardiac surgery. Circulation. 2004;110(suppl):II-237–II-242.
15. Harris KM, Strauss CE, Duval S, Unger BT, Kroshus TJ Inampudi S,
Cohen JD, Kapsner C, Boland LL, Eales F, Rohman E, Orlandi QG,
Flavin TF, Kshettry VR, Graham KJ, Hirsch AT, Henry TD. Multidisci-
1918 Circulation November 1, 2011

plinary standardized care for acute aortic dissection: design and initial
outcomes of a regional care model. Circ Cardiovasc Qual Outcomes.
2010;3:424 – 430.
16. Clouse WD, Hallett JW, Schaff HV, Spittell PC, Rowland CM, Ilstrup
DM, Melton JL. Acute aortic dissection: population-based incidence
compared with degenerative aortic aneurysm rupture. Mayo Clin Proc.
2004;79:176 –180.
17. Olsson C, Thelin S, Stahle E, Ekbom A, Granath F. Thoracic aortic
aneurysm and dissection: increasing prevalence and improved outcomes
reported in a nationwide population-based study of more than 14000
cases from 1987 to 2002. Circulation. 2006;114:2611–2618.
18. Biagini E, Lofiego C, Ferlito M, Fattori R, Rocchi G, Graziosi M, Lovato
L, di Diodoro L, Cooke RM, Petracci E, Bacchi-Reggiani L, Zannoli R,
Branzi A, Rapezzi C. Frequency, determinants, and clinical relevance of
acute coronary syndrome-like electrocardiographic findings in patients
with acute aortic syndrome. Am J Cardiol. 2007;100:1013–1019.
19. Hirata K, Kyushima M, Asato H. Electrocardiographic abnormalities in
patients with acute aortic dissection. Am J Cardiol. 1995;76:1207–1212.
20. Suzuki T, Distante A, Zizza A, Trimarchi S, Villani M, Salerno Uriarte
JA, De Luca Tupputi Schinosa L, Renzulli A, Sabino F, Nowak R,
Birkhahn R, Hollander JE, Counselman F, Vijayendran R, Bossone E,
Eagle K; IRAD-Bio Investigators. Diagnosis of acute aortic dissection
by D-dimer: the International Registry of Acute Aortic Dissection
Substudy on Biomarkers (IRAD-Bio) experience. Circulation. 2009;
119:2702–2707.
21. Shiga T, Wajima Z, Apfel CC, Inoue T, Ohe Y. Diagnostic accuracy of
transesophageal echocardiography, helical computed tomography, and
magnetic resonance imaging for suspected thoracic aortic dissection. Arch
Intern Med. 2006;166:1350 –1356.
22. Pelzel JM, Braverman AC, Hirsch AT, Harris KM. International hetero-
geneity in diagnostic frequency and clinical outcomes of ascending aortic
intramural hematoma. J Am Soc Echocardiogr. 2007;20:1260 –1268.
23. Evangelista A, Mukherjee D, Mehta RH, O’Gara PT, Fattori R, Cooper
JV, Smith DE, Oh JK, Hutchison S, Sechtem U, Isselbacher EM,
Nienaber CA, Pape LA, Eagle KA; International Registry of Aortic
Dissection (IRAD) Investigators. Acute intramural hematoma of the aor-
ta: a mystery in evolution. Circulation. 2005;111:1063–1070.
24. Davies RR, Coe MP, Mandapati D, Gallo A, Botta DM, Elefteriades JA,
Coady MA. Thoracic Surgery Directors Association Award: what is the
optimal management of late-presenting survivors of acute type A aortic
dissection? Ann Thorac Surg. 2007;83:1593–1601.
25. Bavaria JE, Pochettino A, Brinster DR, Gorman RC, McGarvey ML,
Gorman JH, Escherich A, Gardner TJ. New paradigms and improved
results for the surgical treatment of acute type A dissection. Ann Surg.
2001;234:336 –342.

CLINICAL PERSPECTIVE
An acute aortic dissection is a surgical emergency with a high mortality if left untreated. Given the varied presentations,
including similarity to the far more common acute coronary syndromes, diagnosis and appropriate treatment are often
delayed. This report evaluates the reasons for delay in the diagnosis of 894 patients in the International Registry of Acute
Aortic Dissection Registry. Patients with the most typical presenting signs and symptoms, including abrupt onset of severe
chest pain, and those with pulse deficits or hypotension, were diagnosed more quickly. In contrast, patients transferred from
referral hospitals had significantly longer times to diagnosis and ultimately to surgery, perhaps related to the physician’s
experience with dissection at these hospitals. Delays from diagnosis to surgery also occurred in nonwhites, those with prior
cardiac surgery, and those without ongoing shock or hypotension. Education directed at recognition of both typical and
atypical presentations of aortic dissection, particularly to those centers with low exposure to aortic emergencies, may be
of benefit. The fact that the median times from presentation to diagnosis and from diagnosis to surgery exceeded 4 hours
suggests that there is substantial room for improvement. The development of systematic care pathways for diagnosis and
management of aortic dissection, similar to those in place for acute myocardial infarction, may be of benefit. The focus
of these pathways should include recognition of both typical and atypical presentations, with rapid diagnostic imaging in
appropriate candidates and prompt transfer and surgery.